Preparation and characterization of PEG-coated silica nanoparticles for oral insulin delivery.

G Model IJP 14235 No. of Pages 9

International Journal of Pharmaceutics xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm

Pharmaceutical nanotechnology

Preparation and characterization of PEG-coated silica nanoparticles for oral insulin delivery Tatiana Andreani a,b,c , Ana Luiza R. de Souza d , Charlene P. Kiill d , Esteban N. Lorenzón e , Joana F. Fangueiro c,f , Ana Cristina Calpena g , Marco V. Chaud h , Maria L. Garcia i , Maria Palmira D. Gremião d , Amélia M. Silva a,b , Eliana B. Souto c,f,j, * a

Department of Biology and Environment, University of Tras-os Montes e Alto Douro, UTAD, Quinta de Prados, P-5001-801 Vila Real, Portugal Centre for Research and Technology of Agro-Environmental and Biological Sciences (CITAB), UTAD, Vila Real, Portugal Research Centre for Biomedicine (CEBIMED), Fernando Pessoa University (UFP), Praça 9 de Abril, 349, 4249-004 Porto, Portugal d Department of Pharmaceutical Sciences, UNESP – Universidade Estadual Paulista, Rodovia Araraquara-Jau, Km. 01, Araraquara, São Paulo, Brazil e Department of Biochemistry and Chemical Technology, Institute of Chemistry, UNESP – Universidade Estadual Paulista, Araraquara, São Paulo, Brazil f Faculty of Health Sciences, Fernando Pessoa University, Rua Carlos da Maia, 296, P-4200-150 Porto, Portugal g Biopharmacy and Pharmacokicetic Unit, Pharmacy and Pharmaceutical Technology Department, School of Pharmacy, University of Barcelona, Av. Joan XXIII s/n, 8028 Barcelona, Spain h Sorocaba University, UNISO, Rodovia Raposo Tavares, Km. 92.5, 18023-000, Sorocaba, Brazil i Department of Physical Chemistry, Faculty of Pharmacy, Barcelona University, Av. Joan XXIII s/n, 8028 Barcelona, Spain j Institute of Biotechnology and Bioengineering, Centre of Genomics and Biotechnology, UTAD, Vila-Real, Portugal b c

A R T I C L E I N F O

A B S T R A C T

Article history: Received 31 March 2014 Received in revised form 25 July 2014 Accepted 26 July 2014 Available online xxx

The present study reports the production and characterization of PEG-coated silica nanoparticles (SiNP– PEG) containing insulin for oral administration. High (PEG 20,000) and low (PEG 6000) PEG molecular weights were used in the preparations. SiNP were produced by sol–gel technology followed by PEG adsorption and characterized for in vitro release by Franz diffusion cells. In vitro permeation profile was assessed using everted rat intestine. HPLC method has been validated for the determination of insulin released and permeated. Insulin secondary structure was performed by circular dichroism (CD). Uncoated SiNP allowed slower insulin release in comparison to SiNP–PEG. The coating with high molecular weight PEG did not significantly (p > 0.05) alter insulin release. The slow insulin release is attributed to the affinity of insulin for silanol groups at silica surface. Drug release followed second order kinetics for uncoated and SiNP–PEG at pH 2.0. On the other hand, at pH 6.8, the best fitting was first-order for SiNP–PEG, except for SiNP which showed a Boltzmann behavior. Comparing the values of half-live, SiNP–PEG 20,000 showed a faster diffusion followed by Si-PEG 6000 and SiNP. CD studies showed no conformational changes occurring after protein release from the nanoparticles under gastrointestinal simulated conditions. ã 2014 Elsevier B.V. All rights reserved.

Keywords: Silica nanoparticles Insulin PEG adsorption Oral delivery Mathematic modeling HPLC validation

1. Introduction Since the discovery of insulin in 1922 (Banting and Best, 1922), several efforts have been made to achieve the best administration for the treatment of insulin-dependent diabetic patients. Among the numerous advances in this field, oral delivery is considered the most and convenient route for insulin administration because it

* Corresponding author at: Faculty of Health Sciences of Fernando Pessoa University, Rua Carlos da Maia, 296, Office S.1, P-4200-150 Porto, Portugal. Tel.: +351 22 507 4630x3056; fax: +351 22 550 4637. E-mail addresses: [email protected], [email protected] (E.B. Souto).

can overcome the difficulties associated to subcutaneous route such as, daily injections, pain, risk of infection, edema, and fat deposition on the local of injection (Lin et al., 2005). Also, oral insulin delivery can lead to first hepatic bypass which could avoid the hypoglycemic effect that is in many cases a consequence of the subcutaneous insulin administration (Owens et al., 2003). However, the intestinal absorption of pharmaceutical proteins is still a challenge due to their low permeability through the intestinal epithelium as a result of high molecular weight, poor lipophilicity, and surface charge (Lee and Yamamoto, 1990). The association of hydrophilic macromolecules to nanoscale colloidal particles has shown its efficacy to improve protein absorption through the intestinal epithelium resulting in a

http://dx.doi.org/10.1016/j.ijpharm.2014.07.049 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: Andreani, T., et al., Preparation and characterization of PEG-coated silica nanoparticles for oral insulin delivery. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.07.049


2

T. Andreani et al. / International Journal of Pharmaceutics xxx (2014) xxx–xxx

relatively high oral bioavailability (Jung et al., 2000; Lowe and Temple, 1994). Different transport mechanisms have been postulated for explaining the absorption of nanoparticles. They include a paracellular pathway (Bravo-Osuna et al., 2008), transcellular transport (Song et al., 2008) and Peyer’s patches taken up (Borges et al., 2006). Strategies using mucoadhesive delivery systems based on multifunctional polymers may provide a route to increase the bioadhesion to the intestinal mucosa by electrostatic interactions or by the penetration of polymer chains in the mucosa (Llabot et al., 2011; Peppas, 1998; Vllasaliu et al., 2010), thus, improving the transmucosal transport of drugs, specially of peptides and proteins (Janes et al., 2001). Silica nanoparticles (SiNP) are an excellent candidate for the conjugation with biomolecules in the sense that (i) SiNP present large surface area which allows high interaction with drugs (Salonen et al., 2008), (ii) SiNP can act as drug reservoir by possessing high porosity allowing efficient drug loading (Chen et al., 2004), (iii) silica material is known to be biocompatible in vivo application (Arruebo et al., 2006; Dormer et al., 2005), (iv) SiNP are no subjected to microbial attack (Weetal, 1970), (v) these materials can be synthesized at low temperature, which is compatible with the manipulation of biomolecules (Kneuer et al., 2000), and (vi) SiNP possess residual silanol groups (Si– OH) at their surface which can be functionalized by different organic groups (Westcott et al., 1998). Because of its hydrophilic nature, poly(ethylene glycol) (PEG) has been of great interest for the delivery and targeting of various peptides and proteins of therapeutic value, including insulin (Calceti et al., 2004; Elvassore et al., 2001). The addition of PEG to the surface of nanoparticles may allow stabilizing the nanoparticles in the gastric and intestinal fluids due to the inhibition of plasma protein adsorption, and can improve the transport of drugs across the intestinal epithelium (Peppas, 1998; Tobío et al., 2000). Formulated PEG-coating layer around nanoparticles modulated their interaction with intestinal mucosa (Yoncheva etal., 2005), increasing the bioavailability of therapeutic proteins and peptides. Taking into account the ability of PEG to transport drugs across the intestinal epithelium, the main purpose of this work was to study SiNP and PEG–SiNP synthesized by sol–gel technology as carrier for the oral administration of insulin. The nanoparticles were characterized for size distribution by dynamic light scattering (DLS), morphology by atomic force microscopy (AFM) and the structural integrity of insulin after in vitro release studies was analyzed by circular dichroism (CD). The influence of nanoparticles on permeation through everted rat intestine was also evaluated.

under high shear homogenization (Ultra-Turrax, IKA, T25) using NH3 as catalytic agent. The obtained nanoparticles were centrifuged and washed with a mixture of ethanol and ultra-purified water (1:1, v/v) by 2 cycles at 12,000 rpm for 5 min (Spectrafuge16 M, Lambnet International, Inc.). To produce insulin associated to SiNP, 1.0 mL of human insulin (100 UI/mL) was added to 10 mg of uncoated nanoparticles under gentle stirring (300 rpm) for 30 min into ice bath. For insulin–PEG coated-SiNP (Ins–PEG–SiNP), insulin (1.0 mL) was mixed with PEG solutions (2%, w/v) for 30 min at pH 6.8 followed by the addition of SiNP (10 mg) under gentle stirring (300 rpm) for more 30 min into ice bath. The nanoparticles were centrifuged at 5000 rpm for 5 min and the pellet was freeze-dried during 24 h with trehalose (10%, w/v) to prevent particle aggregation.

2. Materials and methods

2.3.3. Morphological studies Nanoparticles were examined morphologically using an atomic force microscopy (AFM). AFM experiments were performed with a Multimode microscope (Veeco, Santa Barbara, California) controlled by Nanoscope IIIa electronics (Veeco). The images were done in tapping mode at 25 C. Samples were prepared by depositing 5 mL of the samples after re-dispersion in ultra-purified water on silicon substrate followed by drying overnight at 25 C.

2.1. Materials Tetraethyl orthosilicate (TEOS, 98%), NH3 25%, PEG with Mw of 6000 and 20,000 Da (PEG 6000; PEG 20,000) were purchased from Merck (Darmstadt, Germany). D-(+)-Trehalose dehydrate was purchased from Sigma–Aldrich (Steinhein, Germany). Solution of 100 U/mL of human insulin (Humulin1 R) was purchased from Eli Lilly (Lisbon, Portugal). Pentobarbital sodium was purchased from Abbot (Rio de Janeiro, Brazil). Ultra-purified water was obtained from Milli1 Q Plus system (Millipore, Germany). All other reagents were of analytical grade. 2.2. Synthesis of nanoparticles SiNP were produced applying the sol–gel technology at room temperature via hydrolysis and condensation of precursor TEOS

2.3. Characterization of nanoparticles 2.3.1. Size and zeta potential analysis The average hydrodynamic diameter (Z-Ave) of SiNP was determined through dynamic light scattering (DLS, Zetasizer Nano ZS, Malvern Instruments, Malvern, UK). DLS is also known as photon correlation spectroscopy (PCS) that allows the measurement of molecules and/or particles due to Brownian motion in submicron region (Goldburg, 1999). Each sample was dissolved with ultra-purified water by vortex before measuring the Z-Ave and polydispersity index (PI). Zeta potential (ZP) of nanoparticles was also measured in purified water adjusting conductivity (50 mS/ cm) with sodium chloride solution (0.9%, w/v), in order to avoid ionic effects upon measurements. The ZP was calculated from the electrophoretic mobility using the Helmholtz–Smoluchowski equation (Deshiikan and Papadopoulos, 1998). Values reported are the mean SD of three different measurement of each sample. 2.3.2. Association efficacy (AE) The association of insulin in nanoparticles was assessed in triplicate using Bradford method. Thirty milligrams of lyophilized nanoparticles were incubated in 3 mL of PBS at pH 6.8 without enzymes under magnetic stirring (150 rpm) for 30 min followed by centrifugation (5000 rpm) for 5 min. The supernatant containing released insulin was collected for protein quantification using a spectrophotometer (Genesys 10 S UV–vis Thermo Scientific) at 595 nm. The amount of insulin associated was calculated according to the following equation: AE ð%Þ ¼

Total amount of insulin insulin in supernatant 100 Total amount of insulin

2.3.4. Apparatus and chromatographic system The system consisted of a waters 1525 pump (Waters, Milford, Massachusetts) with a UV–vis 2487 detector (Waters1). C18 column (5 mm, 250 mm 4.6 mm, Phenomenex1) with a flow rate of 1.0 mL/min was used. The mobile phase consisted of acetonitrile–water with 0.1% TFA (40:60, v:v). The injection volume was 100 mL and a total run time of 12 min. The absorbance of insulin was recorded at 220 nm, and the area under the curve




was measured for calculation of insulin concentration based on the calibration curves. A primary stock solution of insulin was first prepared in water, and subsequently diluted to the appropriated concentration (2, 5, 10, 25, 50, 75, and 100 mg mL1) in different buffers (HCl/KCl, PBS or TC-199) and stored at 4 C for HPLC analysis. Each solution was prepared in triplicate. 2.3.5. Validation of HPLC method A validation for the insulin quantification and permeation test was then required and undertaken in agreement with International Conference on Harmonization guidelines (ICH, 1996) using the following analytical parameters: linearity, precision, accuracy, and specificity. The linearity of method was defined by the slopes, Y-intercept of the calibration curve and the correlation coefficients. The mean peak areas versus concentration of each standard were treated by least-squares linear regression analysis. All calibration curves were obtained from seven different concentrations analyzed in three independent experiments. The specificity was confirmed by comparing placebo samples (buffers without insulin) with insulin to determine possible interferences. Precision intra-assay was evaluated by performing repeatability. For this test, two concentrations (medium and high of the curve) were injected three times in the same day. Accuracy was tested by percentage of the systematic error, which is determined by the relative standard deviation (RSD) 2.3.6. In vitro insulin release and modeling In vitro release studies of insulin from uncoated and coated SiNP were performed using Franz glass diffusion cells. These cells consisted of donor and receptor chambers between which a cellulose nitrate membrane with an average pore size of 0.2 mm is positioned. Phosphate-buffered saline (PBS) at pH 6.8 or HCl/KCl buffer at pH 2.0 are used as receptor fluid and maintained at 37 C. A volume of 300 mL of nanoparticles (15 mg) was applied to the donor compartment containing 7 mL of buffer. During all the experiment, a magnetic bar was stirring in each cell. At determined times, 1 mL of the samples were collected and the same volume was replaced with buffer. The insulin collected was analyzed by HPLC. The insulin release data were also kinetically evaluated applying the Akaike’s approach (Doménech et al., 1998). The empirical models were selected based on the Akaike Information Criteria (AIC). The model associated to the smallest value of AIC is considered as giving the best fit of the set of data. The AIC was calculated according to the following equation: AIC ¼ n ln SSQ þ 2p where n is the number of pairs of experimental values, SSQ is the residual sum of squares and p is the number of parameters of the model. The mean release time (MRT) of insulin from different nanoparticles was also arithmetically assessed from experimental data using the equation: MRT ¼

S t i DQ Q1

where Q1 is the asymptote of the dissolved amount of drug and Sti DQ is the area between the cumulative release and Q1. 2.3.7. Circular dichroism The secondary structure of insulin was determined by circular dichroism (CD). The CD spectra of native insulin and insulin released into buffers at pH 2.0 and pH 6.8 from nanoparticles were

3

recorded in a 0.1 cm cell from 250 to 190 nm at 25 C using a Jasco Spectropolarimeter J-815 (Tokyo, Japan). CD spectra of all samples were acquired and subtracted from the reference spectra (polymer solution in the respective pH). Each spectrum represents an average of 6 successive scans and is expressed as the mean residual ellipticity [u] (deg cm2 dmol1). Deconvolution of CD spectra was obtained by the SELCON3 method (Sreerama and Woody, 2000). 2.3.8. Permeation studies through everted rat intestine Male Wistar rats, weighing approximately 250–270 g were provided by São Carlos Federal University (Animal Ethics Committee, protocol no. 003/2011). The sample tested were insulin-loaded uncoated and PEGylated nanoparticles dissolved in TC-199 buffer solution with lactose. The TC-199 buffer is composed of NaCl (145 mM), KCl (4.56 mM), CaCl22H2O (1.25 mM), NaHPO4 (5 mM) and lactose (10 mM). Rats fasted overnight with free access to water were anesthetized by an intraperitoneal injection of sodium pentobarbital (0.5 mL/kg) and killed by cervical dislocation immediately after removing the small intestine via an abdominal incision. Small intestine (ca. 6 cm) was removed, rinsed several times with oxygenated TC-199 buffer solution without lactose at 10 C and gently everted. One end of the gut sac was clamped and the whole length of the intestine was filled with TC-199 buffer and sealed with braided silk sutures. The everted gut sacs were placed into Erlenmeyer flasks containing a solution of TC-199 buffer with lactose (20 mL) and nanoparticles (loaded with insulin), and incubated at 37 C. The solution was continually aerated. After 60 min, the sacs were removed, washed with TC-199 and cut open. The content of the sacs were then filtered via cellulose acetate membrane (0.22 mm). The cumulative amount of insulin permeated through the gut sac was analyzed by HPLC. 2.3.9. Statistical analysis Statistically evaluation of data was performed using a one-way analysis of variance (ANOVA) test. Bonferroni’s multiple comparison test was carried out to compare the significance between the different groups. A p-value < 0.05 was considered statistically significant. 3. Results and discussion 3.1. Size and zeta potential analysis In order to develop an effective mucoadhesive vehicle for oral insulin administration, formulations were synthesized by coating SiNP with two different PEG molecules (6000 or 20,000 Da). The effect of PEG coating is known to increase protein stability and promote specific interactions with mucosal tissues (Tobio et al., 1998; Vila et al., 2005). In the development of oral drug delivery systems, parameters such as the particle size may influence their clearance by the gastrointestinal tissue and the bioadhesion to the biological cells (Desai et al., 1996). The measured Z-Ave PI, ZP, and AE of Ins–SiNP prepared in the absence and presence of PEG with different molecular weights are summarized in Table 1. The synthesis of nanoparticles by sol–gel technology under mild conditions resulted in SiNP with an average diameter of 289.60 28.24 nm, with a PI of 0.251 0.081. The average hydrodynamic diameter of the particles increased after the PEGylation, as well as the PI. This tendency was more pronounced for PEG 20,000 showing Z-Ave of 625.20 20.21 nm and PI of 0.315 0.030 than for PEG 6000 that showed in a Z-Ave of 493.7 89.10 nm and PI of 0.580 0.010. Such observations can be due to an increase of the extending PEG layer coating the final particles. The molecular weight of PEG influences the size and size



4


Table 1 The results of Z-Ave PI, ZP, and AE of the Ins-nanoparticles prepared by sol–gel technology (mean SD, n = 3).

Ins–SiNP Ins–SiNP–PEG 6000 Ins–SiNP–PEG 20,000

Z-Ave SD (nm)

PI SD

ZP SD (mV)

Insulin AE SD (%)

289.6 28.24 493.7 89.10 625.2 20.21

0.251 0.081 0.580 0.010 0.315 0.030

25.0 0.1 15.2 0.0 13.8 0.1

71.7 85.4 82.5

distribution, as well as the ZP of nanoparticles. With increasing the chain length of PEG, the particle size tends to increase. The PI also increases, possibly due to nanoparticle aggregation, following the presence of PEG chains at silica nanoparticle surface. After coating with PEG, the nanoparticles tend to aggregate, due to the high density of PEG which results in coalescence, or due to the presence of small and large fragments by the input of energy during nanoparticle synthesis. The ZP values of nanoparticles varied from 25 0.1 mV to 13.8 0.1 mV. The high negative ZP value of SiPN is attributed to the presence of deprotonated silanol molecules on the silica nanoparticle surface. The results presented in Table 1 also show that the ZP was affected by the MW of the polymer used for synthesizing the nanoparticles. In addition, the adsorption of PEG influences the distribution of charge in the diffuse part of the electrical double layer, reducing the ZP. These results can be due to the coating of silanol groups, which were used as reactive site for PEG adsorption (Prokopowicz and Łukasiak, 2010). Since PEG 20,000 has high MW, the interaction with silica nanoparticles is more pronounced than with PEG 6000 and, thus the ZP is decreased. It is known that particle uptake is strongly influenced by surface charge of particles. Due to the presence of negative charges in the mucin, opposite electric charges are required to increase the residence time of particles and consequently to achieve better drug absorption (George and Abraham, 2006). However, high interaction between positively charged nanoparticles and mucin can lead to a decrease in their absorption and consequently their uptake by intestinal mucosa. In addition, some studies have reported that the use of anionic polymers leads to better mucosal adhesion than cationic polymers or non-ionic polymers (Chickering and

Mathiowitz, 1995). This observation can be attributed to the presence of numerous surface carboxyl groups on anionic polymers, generating strong biodhesive interactions by hydrogen bonds with the mucin. Therefore, the negative charge of prepared nanoparticles is expected to improve the protein absorption. 3.2. Association efficacy It is known that by modifying the surface characteristics of nanoparticles it is possible to obtain different values of drug efficiency association (AE). The EA values with insulin to SiNP and to PEG-coated SiNP were of 71.7, 85.4, and 82.5% for SiNP, SiNP–PEG 6000, and SiNP–PEG 20,000, respectively (Table 1). During PEG coatings, the value of AE for PEG-coated nanoparticles increases by the adsorption of insulin on the surface of particles attributed to the affinity of the protein to the hydrophilic PEG chains. 3.3. Morphological studies AFM studies were conducted to investigate the morphology of prepared nanoparticles (Fig. 1). AFM observations showed individualized spherical-shaped SiNP with minor deformation and aggregation (Fig. 1, left). Coating with PEG and the insulin association with SiNP did not affect the particle shape. However, from the images, partial aggregation of nanoparticles was observed after coating with PEG 6000 (Fig. 1, right). Also, SiNP have a smaller size in comparison to that observed by DLS characterization, probably due to the shrinkage and collapse of SiNP and PEG layer in the drying process. In all nanoparticles the particle surface is very smooth and no signal of insulin is visible.

Fig. 1. Morphological characterization of SiNP (left) and Ins-SiNP coated with PEG 6000 (right).




5

D

A

0.20

0.10

0.15 0.10 AU

AU

0.05 0.00

0.05 0.00

-0.05

-0.05

-0.10 1.00

2.00

3.00

4.00

5.00 Minutes

6.00

7.00

8.00

9.00

-0.10

10.00

B

E

0.15

0.10

2.00

3.00

4.00

5.00 Minutes

6.00

7.00

8.00

9.00

10.00

0.05

0.05

AU

AU

0.10

1.00

0.00

0.00

-0.05

-0.05 -0.10

-0.10 1.00

2.00

3.00

4.00

5.00 Minutes

6.00

7.00

8.00

9.00

10.00

C

1.00

2.00

3.00

4.00

5.00 Minutes

6.00

7.00

8.00

9.00

10.00

1.00

2.00

3.00

4.00

5.00 Minutes

6.00

7.00

8.00

9.00

10.00

F

0.15

0.10

0.10

0.05 AU

AU

0.05 0.00 -0.05

0.00 -0.05

-0.10 1.00

2.00

3.00

4.00

5.00 Minutes

6.00

7.00

8.00

9.00

10.00

-0.10

Fig. 2. Representative chromatograms after application of the buffers without insulin HCl/KCl buffer (D), PBS (B), TC-199 (C) and after standard solution of 1.96 mg/mL insulin in HCl/KCl (A), PBS (E) and TC-199 (F), respectively.

3.4. Validation of HPLC method The quantitative determinations of insulin for the validation of release and permeation assays require specific and sensitive methods. Calibration curves from seven human insulin standard solutions of different concentrations ranging from 1.96 to 98 mg mL1 were prepared with HCl/KCl buffer pH 2.0, PBS pH 6.8 or TC199 (without lactose) pH 7.2. A good linearity was obtained over the insulin concentration range of proposed range. The representative linear equations were y = 212957.35x 92104.14, y = 213123.34x 103899.26, and y = 200583.62x 43289.48 for HCl/KCl, PBS and TC 199 (without lactose), respectively. All the

correlation coefficients (r) were found to be 0.9999, indicating a good linearity in the range of study. Specificity of method was tested by analyzing potential interfering peaks of the buffers at insulin retention time. Three injection of buffer solutions were performed in the same conditions as described for insulin solution injection. The method was found to be specific since no interfering peaks were observed as shown in Fig. 2. Precision of the method was determined by analyzing the samples at two different concentrations. For the intra-day analysis, samples were analyzed in triplicate (n = 3). The peak area RSD (%) values for the concentration of 24.5 mg/mL were 2.45, 3.45, and

Table 2 Validation parameters for insulin quantification in HCl/KCl buffer (pH 2.0). Theorical concentration (mg/mL)

Experimental concentration (mg/mL)

Standard deviation (SD)

Recovery (%)

RSD (%)

1.96 4.90 9.80 24.5 49.0 73.5 98.0

1.99 5.20 9.40 24.23 48.31 73.51 98.16

0.07 0.11 0.12 0.32 0.51 1.93 3.02

101.53 106.12 95.91 98.89 98.59 100.00 100.16

3.55 2.22 1.34 1.34 1.05 2.62 3.08



6


Table 3 Validation parameters for insulin quantification in PBS (pH 6.8). Theorical concentration (mg/mL)



Recovery (%)

RSD (%)

1.96 4.90 9.80 24.5 49.0 73.5 98.0

1.63 4.52 8.73 23.56 48.14 72.84 97.48

0.10 0.09 0.10 0.26 0.42 1.57 2.46

83.16 92.24 89.08 96.16 98.24 99.10 99.46

6.14 2.08 1.18 1.12 0.87 2.16 2.53

Table 4 Validation parameters for insulin quantification in TC-199 buffer (pH 7.2). Theorical concentration (mg/mL)



Recovery (%)

RSD (%)

1.96 4.90 9.80 24.5 49.0 73.5 98.0

1.74 4.91 9.97 24.30 48.91 73.54 98.09

0.05 0.03 0.39 0.57 0.55 0.64 1.79

88.94 100.34 101.74 99.20 99.82 100.06 100.09

3.28 0.65 1.79 1.47 0.68 0.58 1.14

Fig. 3. Cumulative release profiles of insulin from different nanoparticles. Experiments were conducted in HCl/KCl buffer at pH 2.0 (A) and in PBS at pH 6.8 (B) (n = 6 SD).

0.42 for HCl/KCl buffer, PBS and TC-199 buffer, respectively. For 98 mg/mL the RSD (%) values were 1.10, 0.17, and 0.73 for HCl/KCl buffer, PBS and TC-199 buffer, respectively. Accuracy of the method was defined as the percentage recoveries of mean of a known amount of insulin, which is assessed as standardized agreement between the measured value and the true value by determination of the RSD. The recovery varied from 91.91 to 106.12% for HCl/KCl buffer, from 83.16 to 99.46% for PBS and from 88.94 to 102.74% for TC-199 buffer, respectively, satisfying the acceptance criteria (Tables 2–4).

3.5. In vitro insulin release and modeling The insulin release profile from uncoated and coated SiNP in HCl/KCl buffer or PBS at 37 C is presented in Fig. 3A and B. Experimental results showed that at both pH conditions (pH 2.0, pH 6.8), the release of insulin was delayed from uncoated and PEG-coated SiNP in comparison to insulin solution (Fig. 3A and B). At both pH, only SiNP and SiNP–PEG 6000 affected significantly the insulin release profile (p < 0.05). Insulin-loaded SiNP synthesized in the absence of PEG displayed lower rate of release drug from colloidal system than

Table 5 Mathematical modeling for insulin release at gastric and intestinal conditions. Release condition

Gastric

Intestinal

Formulations

Insulin free Ins–SiNP Ins–SiNP–PEG Ins–SiNP–PEG Insulin free Ins–SiNP Ins–SiNP–PEG Ins–SiNP–PEG

Zero-order

20,000 6000

20,000 6000

First-order

Second order

Boltzmann

Hiperbola

AIC

r2

AIC

r2

AIC

r2

AIC

r2

AIC

39.86 31.45 36.90 35.42 61.09 54.21 58.73 54.75

0.9738 0.9929 0.9767 0.9798 0.9427 0.9761 0.9492 0.9570

18.29 29.58 20.05 16.27 29.70 35.06 26.12 12.06

0.9982 0.9944 0.9972 0.9982 0.9986 0.9973 0.9989 0.9997

19.02 13.01 12.43 9.09 52.51 29.70 48.44 42.60

0.9985 0.9994 0.9991 0.9994 0.9801 0.9983 0.9852 0.9896

* 14.94 16.57 14.79 28.41 25.75 29.00 *

* 0.9999 0.9993 0.9999 0.9992 0.9993 0.9990 *

81.51 102.75 79.89 67.85 39.04 39.17 34.81 22.01




7

Table 6 MRT and half-life for In–SiNP, Ins–SiNP–PEG 6000, and Ins–SiNP–PEG 20,000 at gastric pH. Gastric

Insulin solution

Ins–SiNP

Ins–SiNP–PEG 6000

Ins–SiNP–PEG 20,000

MRT (min) Half-live (min)

34.10 0.9 30.32

52.80 0.3 194.90

36.2 1.2 35.05

35.8 1.3 34.15

The statistically different formulations are: insulin solution vs. Ins–SiNP: p < 0.0001; Ins–SiNP vs. Ins-SiNP–PEG 20,000: p < 0.0001; Ins–SiNP vs. Ins–SiNP–PEG 6000: p < 0.0001; insulin solution vs. Ins–SiNP–PEG 6000: p < 0.05; insulin solution vs. Ins–SiNP–PEG 20,000: p < 0.05. Table 7 MRT and half-life for insulin, Ins–SiNP, Ins–SiNP–PEG 6000, and Ins–SiNP–PEG 20,000 at intestinal pH. Intestinal

Insulin solution

Ins–SiNP

Ins–SiNP–PEG 6000

Ins–SiNP–PEG 20,000

MRT (min) Half-life (min)

50.0 0.4 33.45

86.6 2.0 89.24

57.5 2.1 36.16

57.5 2.1 36.01

The statistically different formulations are: insulin solution vs. Ins–SiNP: p < 0.0001; Ins–SiNP vs. Ins–SiNP–PEG 20,000: p < 0.0001; Ins–SiNP vs. Ins–SiNP–PEG 6000: p < 0.0001; insulin solution vs. Ins–SiNP–PEG 6000: p < 0.001; insulin solution vs. Ins–SiNP–PEG 20,000: p < 0.05; Ins–SiNP–PEG 6000 vs. Ins–SiNP–PEG 20,000: p < 0.01.

A 5000

2

[θ] deg cm dmol

-1

0 -5000 -10000 -15000 -20000 -25000 -30000 200

210

220

230

240

250

240

250

Wavelength (nm)

B 5000

-5000

2

[θ] deg cm dmol

-1

0

-10000 -15000 -20000 -25000 200

210

220

230

Wavelenght (nm) Fig. 4. Far-UV CD spectrum of insulin solution (black line), insulin released from SiNP–PEG 6000 (black dotted line) and SiNP–PEG 20,000 (grey line) at pH 2.0 (A) and pH 6.8 (B) buffers.

of PEG-coated SiNP at both pH. This result can be best described by the fact that insulin could be easily bound to silanol groups present onto the silica surface leading to a lower insulin release rate. With

the addition of PEG, the number of sites for insulin association decreased, resulting in faster protein release. In addition, PEG coating can lead to an increase of drug release due to the hydration process of PEG layers in the dispersion medium which favors the protein diffusion (Quellec et al., 1998). The results also show that the released amount depended on the insulin associated to nanoparticles. Higher insulin association resulted in more insulin release, which is due to the higher insulin adsorption on the surface of nanoparticles, leading to a faster release rate in the initial phase. The retention time of particles in stomach is considered to be 1–2 h. Therefore, the amount of insulin release at pH 2.0 (60%), after 2 h, indicates that it is important to optimize the system in order to provide a minimum contact between protein and gastric environment, and, thus, avoiding a possible protein denaturation. A different situation arises from studies with buffer at pH 6.8 simulating intestinal conditions for insulin released from uncoated SiNP (Fig. 3B). When PBS was used, lower insulin release from SiNP was observed in the first 1 h than in the acidic pH. This result can be attributed to the fact that low pH, disfavors the polycondensation reactions which results in faster dissolution of silica matrix (Kortesuo et al., 2002). On the other hand, no significant differences were found comparing the insulin release from PEGylated SiNP at pH 2.0 and 6.8. Insulin release from different nanoparticles was kinetically studied using zero-order, first-order, second-order, Boltzmann, and hiperbola models. Table 5 lists the models and their corresponding dissolution parameters for insulin and different nanoparticles at pH 2.0 and pH 6.8. According to lower AIC values, at pH 2.0, for insulin solution, the best model was found to be first order, and for SiNP and SiNP coated with PEG 6000 or PEG 20,000 the best model was found to be second order. At intestinal condition (pH 6.8) the best fitting was first-order for all formulation, except for SiNP and insulin solution, which showed a Boltzmann behavior. Comparing the values of half-lives, at both acidic and neutral pH, insulin solution showed a slower half-live resulting in slower time for insulin release (Tables 6 and 7). Insulin solution showed a faster diffusion followed by SiNP–PEG 20,000, Si–PEG 6000, and SiNP. Therefore the association of insulin to SiNP considerably prolongs its half-life. Tables 6 and 7 also show the MRT for insulin and different nanoparticles at both gastric and intestinal pH, respectively. The MRT obtained with uncoated and coated silica nanoparticles increases compared with insulin solution, confirming that the association of insulin to nanoparticles modified its release time.



8


Table 8 Observed parameters of secondary insulin solution structure and of insulin release from SiNP–PEG 6000 and SiNP–PEG 20,000 at gastric and at intestinal conditions. Condition

Samples

a-Helix (%)

b-Sheet (%)

b-Turn (%)

Random coil (%)

Gastric

Insulin solution SiNP–PEG 20,000 SiNP–PEG 6000 Insulin solution SiNP–PEG 20,000 SiNP–PEG 6000

66.8 59.1 53.6 66.8 50.0 39.7

1.7 3.6 4.7 1.7 6.6 15.6

10.2 14.9 14.0 10.2 17.5 15.5

23.4 23.7 30.6 23.4 29.0 29.3

Intestinal

3.6. Circular dichroism

4. Conclusions

Circular dichroism (CD) is an excellent tool for investigating the secondary structure of proteins in solution. The far-UV range of the CD spectra provides information about a-helix and b-sheet conformation on various protein structures. The negative band at 223 nm is based on the anti-parallel b-structure of insulin and is mainly due to its monomeric form. In contrast, the negative band at 208 nm is assigned to its a-helical structure and is due to its dimer (Pocker and Biswas, 1980; Woody, 1996). The structure of insulin was studied after release from nanoparticles at pH 2.0 and pH 6.8 using CD spectroscopy (Fig. 4A and B). CD spectra of insulin released from all nanoparticles at pH 2.0 and pH 6.8 revealed a slight difference compared with native insulin. Parameters of the secondary structure of insulin, including a-helix, b-sheet, b-turn, and random coil, are summarized in Table 8 at pH 2.0 and pH 6.8. At both pH, some differences can be seen comparing the secondary structure of insulin solution with insulin release from SiNP–PEG 6000 and SiNP–PEG 20,000. The content of a-helix decreases, while the content of b-sheet increases. The content of b-turn and random coil does not obviously change. These results can be attributed to the adsorption of insulin on the surface of particles because of the affinity of the protein with the hydrophilic PEG chains, as confirmed by high AE. The association of insulin to hydrophilic polymers could make the a-helix draw and extend into b-sheet (Liu et al., 2007). Regardless of slight difference of the spectra intensity and fractional composition of secondary parameters, the line shape of the spectrum of insulin released was similar to that of the native insulin, indicating that the association of insulin to nanoparticles to did not influence the conformation of the protein.

Diabetes is a chronic disease with high clinical significance. Therefore, an urgent option for overcome the difficulties associated to the subcutaneous injection is required. The oral administration represents the most convenient and less invasive route for insulin delivery. Of the various new drug delivery systems for the therapeutic treatment of diabetes, colloidal nanoparticle carriers seem to be a promise strategy for oral insulin administration. According to these results, the incorporation of PEG at silica surface changed their physicochemical properties, influencing the particle size and size distribution, the insulin association efficacy, as well the insulin in vitro release profile. The in vitro study exhibited a modification in the release rate for the nanoparticles depending on the coating of the SiNP used. The PEGylation of SiNP resulted in faster insulin release at gastric and intestinal conditions in comparison to uncoated SiNP due to low interaction between insulin and silanol groups present onto silica surface. However, a further optimization of the release profile is required to decrease the insulin release in the low-pH gastric environment. The result reported in the present work showed that the presence of PEG onto the silica nanoparticles did not increase the permeation behavior of insulin through the small intestinal mucosa. Therefore, further in vivo studies will also be carried out with silica nanoparticles to evaluate the ability of nanoparticles to improve the insulin permeation, as well as the protein bioavailability after associating to nanoparticles.

3.7. Permeation studies through everted rat intestine Permeation studies were carried out to investigate the influence of PEG coatings on insulin transport through the intestinal mucosa. The results of percentage of insulin permeation were 46.50% 4.82, 42.46% 5.08, 48.70% 3.86, and 40.13% 2.77 for insulin solution, SiNP, SiNP–PEG 6000, and SiNP–PEG 20,000 respectively. Although PEG is a polymer that may increase the drug absorption through intestinal mucosa, the presence of PEG onto the silica surface did not significantly change the permeation behavior of insulin through the small intestinal mucosa (p < 0.05) Several studies have demonstrated that the conjugation of low MW PEGs to drugs can enhance the absorption. However, the effect of PEG on protein permeation in intestinal mucosa is strongly associated to the polymer structure, to the molecular weight of PEG, as well as to the percentage of polymer used for coating strategy. Also, in order to avoid the limiting factors of the use of everted rat intestine studies, such as loss of protein, in vivo models should be applied for better results of nanoparticles permeation.

Acknowledgments The work was partially supported by Fundação para a Ciência e Tecnologia (FCT, Portugal), namely, the PhD scholarships SFRH/BD/ 60640/2009 for T. Andreani and SFRH/BD/80335/2011 for J.F. Fangueiro. FCT is also acknowledged under the reference PTDC/ SAU-FAR/113100/2009 and PEst-C/AGR/UI4033/2011. The authors acknowledge the support by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) for the Ph.D. scholarship for C.P. Kiill under the reference 2012/10174-3 and for E.N. Lorenzón under the reference 140758/2011-9. References Arruebo, M., Galan, M., Navascues, N., Tellez, C., Marquina, C., Ibarra, M.R., Santamaria, J., 2006. Development of magnetic nanostructured silica-based materials as potential vectors for drug-delivery applications. Chem. Mater. 18, 1911–1919. Banting, F.G., Best, C.H., 1922. The internal secretion of the pancreas. J. Lab. Clin. Med. 7, 251–266. Borges, O., Cordeiro-da-Silva, A., Romeijn, S.G., Amidi, M., de Sousa, A., Borchard, G., Junginger, H.E., 2006. Uptake studies in rat Peyer’s patches, cytotoxicity and release studies of alginate coated chitosan nanoparticles for mucosal vaccination. J. Control. Release 114, 348–358. Bravo-Osuna, I., Vauthier, C., Chacun, H., Ponchel, G., 2008. Specific permeability modulation of intestinal paracellular pathway by chitosan-poly(isobutylcyanoacrylate) core-shell nanoparticles. Eur. J. Pharm. Biopharm. 69, 436–444. Calceti, P., Salmaso, S., Walker, G., Bernkop-Schnürch, A., 2004. Development and in vivo evaluation of an oral insulin–PEG delivery system. Eur. J. Pharm. Sci. 22, 315–323.



T. Andreani et al. / International Journal of Pharmaceutics xxx (2014) xxx–xxx Chen, J.-F., Ding, H.-M., Wang, J.-X., Shao, L., 2004. Preparation and characterization of porous hollow silica nanoparticles for drug delivery application. Biomaterials 25, 723–727. Chickering, D.E., Mathiowitz, E., 1995. Bioadhesive microspheres: I. A novel electrobalance-based method to study adhesive interactions between individual microspheres and intestinal mucosa. J. Control. Release 34, 251–262. Desai, M., Labhasetwar, V., Amidon, G., Levy, R., 1996. Gastrointestinal uptake of biodegradable microparticles: effect of particle size. Pharm. Res. 13, 1838–1845. Deshiikan, S.R., Papadopoulos, K.D., 1998. Modified booth equation for the calculation of zeta potential. Colloid Polym. Sci. 276, 117–124. Doménech, J., Martínez, J., Plá, J.M., 1998. Biofarmacia y Farmacocinética II. Síntesis SA, Madrid. Dormer, K., Seeney, C., Lewelling, K., Lian, G., Gibson, D., Johnson, M., 2005. Epithelial internalization of superparamagnetic nanoparticles and response to external magnetic field. Biomaterials 26, 2061–2072. Elvassore, N., Bertucco, A., Caliceti, P., 2001. Production of insulin-loaded poly (ethylene glycol)/poly(l-lactide) (PEG/PLA) nanoparticles by gas antisolvent techniques. J. Pharm. Sci. 90, 1628–1636. George, M., Abraham, T.E., 2006. Polyionic hydrocolloids for the intestinal delivery of protein drugs: alginate and chitosan—a review. J. Control. Release 114, 1–14. Goldburg, W.I., 1999. Dynamic light scattering. Am. J. Phys. 67, 1152–1160. ICH, 1996. International Conference on Harmonization (ICH) of Technical Requirements for the Registration of Pharmaceuticals for Human Use Validation of Analytical Procedures: Methodology. (ICB-Q2B). Janes, Calvo, P., Alonso, M.J., 2001. Polysaccharide colloidal particles as delivery systems for macromolecules. Adv. Drug Deliv. Rev. 47, 57–83. Jung, T., Kamm, W., Breitenbach, A., Kaiserling, E., Xiao, J.X., Kissel, T., 2000. Biodegradable nanoparticles for oral delivery of peptides: is there a role for polymers to affect mucosal uptake? Eur. Pharm. Biopharm. 50, 147–160. Kneuer, C., Sameti, M., Haltner, E.G., Schiestel, T., Schirra, H., Schmidt, H., Lehr, C.-M., 2000. Silica nanoparticles modified with aminosilanes as carriers for plasmid DNA. Int. J. Pharm. 196, 257–261. Kortesuo, P., Ahola, M., Kangas, M., Jokinen, M., Leino, T., Vuorilehto, L., Laakso, S., Kiesvaara, J., Yli-Urpo, A., Marvola, M., 2002. Effect of synthesis parameters of the sol–gel processed spray-dried silica gel microparticles on the release rate of dexmedetomidine. Biomaterials 23, 2795–2801. Lee, V.H.L., Yamamoto, A., 1990. Penetration and enzymatic barriers to peptide and protein absorption. Adv. Drug Deliv. Rev. 4, 171–207. Lin, Y.H., Liang, H.F., Chung, C.K., Chen, M.C., Sung, H.W., 2005. Physically crosslinked alginate/N,O-carboxymethyl chitosan hydrogels with calcium for oral delivery of protein drugs. Biomaterials 26, 2105–2113. Liu, T., Guo, R., Hua, W., Qiu, J., 2007. Structure behaviors of hemoglobin in PEG 6000/ Tween 80/Span 80/H2O niosome system. Colloids Surf. B 293, 255–261. Llabot, J.M., Salman, H., Millotti, G., Bernkop-Schnürch, A., Allemandi, D., Manuel Irache, J., 2011. Bioadhesive properties of poly(anhydride) nanoparticles coated with different molecular weights chitosan. J. Microencapsul. 28, 455–463.

9

Lowe, P.J., Temple, C.S., 1994. Calcitonin and insulin in isobutylcyanoacrylate nanoparticles: protection against proteases and effect on intestinal absorption in rats. J. Pharm. Pharmacol. 46, 547–552. Owens, D.R., Zinman, B., Bolli, G., 2003. Alternative routes of insulin delivery. Diabetic Med. 20, 886–898. Peppas, N.A., 1998. Molecular calculations of poly(ethylene glycol) transport across a swollen poly(acrylic acid)/mucin interface. J. Biomater. Sci. Polymer. Edn. 9, 535–542. Pocker, Y., Biswas, S.B., 1980. Conformational dynamics of insulin in solution. Circular dichroic studies. Biochemistry 19, 5043–5049. Prokopowicz, M., Łukasiak, J., 2010. Synthesis andin vitro characterization of freezedried doxorubicin-loaded silica/PEG composite. J. Non-Crystal. Sol. 356, 1711–1720. Quellec, P., Gref, R., Perrin, L., Dellacherie, E., Sommer, F., Verbavatz, J.M., Alonso, M. J., 1998. Protein encapsulation within polyethylene glycol-coated nanospheres. I. Physicochemical characterization. J. Biomed. Mater. Res. 42, 45–54. Salonen, J., Kaukonen, A.M., Hirvonen, J., Lehto, V., 2008. Mesoporous silicon in drug delivery applications. J. Pharm. Sci. 97, 632–653. Song, Q., Yao, L., Huang, M., Hu, Q., Lu, Q., Wu, B., Qi, H., Rong, Z., Jiang, X., Gao, X., Chen, J., Chen, H., 2008. Mechanisms of transcellular transport of wheat germ agglutinin-functionalized polymeric nanoparticles in Caco-2 cells. Biomaterials 33, 6769–6782. Sreerama, N., Woody, R., 2000. Estimation of protein secondary structure from circular dichroism spectra: comparison of CONTIN SELCON and CDSSTR methods with an expanded reference set. Anal. Biochem. 287, 252–260. Tobio, M., Gref, R., Sanchez, A., Langer, R., Alonso, M.J., 1998. Stealth PEG–PLA nanoparticles as protein carriers for nasal administration. Pharm. Res. 15, 270–275. Tobío, M., Sánchez, A., Vila, A., Soriano, I.I., Evora, C., Vila-Jato, J.L., Alonso, M.J., 2000. The role of PEG on the stability in digestive fluids and in vivo fate of PEG–PLA nanoparticles following oral administration. Colloids Surf. B 18, 315–323. Vila, A., Sanchez, A., Évora, C., Soriano, I., Mccallion, O., Alonso, M.J., 2005. PLA–PEG particles as nasal protein carriers: the influence of the particle size. Int. J. Pharm. 292, 43–52. Vllasaliu, D., Exposito-Harris, R., Heras, A., Casettari, L., Garnett, M., Illum, L., Stolnik, S., 2010. Tight junction modulation by chitosan nanoparticles: comparison with chitosan solution. Int. J. Pharm. 400, 183–193. Weetal, H.H., 1970. Starage stability of water insiluble enzymes covalently coupled to organic and inorganic carriers. Biochim. Biophys. Acta 212, 1–7. Westcott, S.L., Oldenburg, S.J., Lee, T.R., Halas, N.J., 1998. Formation and adsorption of clusters of gold nanoparticles onto functionalized silica nanoparticle surfaces. Langmuir 14, 5396–5401. Woody, R.W., 1996. Circular Dichroism and the Conformational Analysis of Biomolcules. Plenum Press, New York. Yoncheva, K., Lizarraga, E., Irache, J.M., 2005. PEGylated nanoparticles based on poly (methyl vinyl ether-co-maleic anhydride): preparation and evaluation of their bioadhesive properties. Eur. J. Pharm. Sci. 24, 411–419.


Synthesis, characterization, and in vitro studies of PLGA-PEG nanoparticles for oral insulin delivery.

CapsMorph® technology for oral delivery--theory, preparation and characterization.

Novel PLGA-based nanoparticles for the oral delivery of insulin.

Preparation and characterization of rifampin loaded mesoporous silica nanoparticles as a potential system for pulmonary drug delivery.

Poly(n-butylcyanoacrylate) nanoparticles for oral delivery of quercetin: preparation, characterization, and pharmacokinetics and biodistribution studies in Wistar rats.

Preparation, characterization, and in vivo study of rhein-loaded poly(lactic-co-glycolic acid) nanoparticles for oral delivery.

Preparation and optimization of triptolide-loaded solid lipid nanoparticles for oral delivery with reduced gastric irritation.

Preparation and characterization of antioxidant nanoparticles composed of chitosan and fucoidan for antibiotics delivery.

Programmable drug release using bioresponsive mesoporous silica nanoparticles for site-specific oral drug delivery.

pH-Responsive Mesoporous Silica and Carbon Nanoparticles for Drug Delivery.

silica nanocomposites: preparation and characterization.

Preparation and characterization of amphiphilic calixarene nanoparticles as delivery carriers for paclitaxel.

Preparation and characterization of anionic pullulan thermoassociative nanoparticles for drug delivery.

Preparation, characterization, and safety evaluation of poly(lactide-co-glycolide) nanoparticles for protein delivery into macrophages.

Preparation and characterization of amorphous amphotericin B nanoparticles for oral administration through liquid antisolvent precipitation.

Preparation and characterization of betulin nanoparticles for oral hypoglycemic drug by antisolvent precipitation.

Characterization and optimization of pH-responsive polymer nanoparticles for drug delivery to oral biofilms.

Elaboration and characterization of barley protein nanoparticles as an oral delivery system for lipophilic bioactive compounds.

Oral insulin-delivery system for diabetes mellitus.

Candesartan cilexetil loaded solid lipid nanoparticles for oral delivery: characterization, pharmacokinetic and pharmacodynamic evaluation.

Preparation of novel curdlan nanoparticles for intracellular siRNA delivery.

Advances in preparation and characterization of chitosan nanoparticles for therapeutics.

Preparation of silica nanoparticles through microwave-assisted acid-catalysis.

Glucose- and temperature-sensitive nanoparticles for insulin delivery.